Skip to contents

DBSCAN clustering, BFS over density-connected cores (hmdbs)

Source: R/geron_w4a_native.R
morie_geron_dbscan.Rd

Core/border/noise classification is delegated to morie_geron_dbscan_core_point; this wrapper merges cores into clusters and assigns border points.

Usage

morie_geron_dbscan(X, eps, min_samples, metric = "euclidean")

Arguments

X

Points (m, d).

eps

Neighbourhood radius (>0).

min_samples

Core threshold (>=1).

metric

"euclidean"/ "manhattan"/"chebyshev".

Value

List with labels, n_clusters, n_noise, n_core, n_border, core_indices, cluster_sizes.

Examples

morie_geron_dbscan(X = c(1, 2, 3, 4, 5, 6, 7, 8), eps = 0.5, min_samples = 5L)
#> $labels
#> [1] -1 -1 -1 -1 -1 -1 -1 -1
#> 
#> $n_clusters
#> [1] 0
#> 
#> $n_noise
#> [1] 8
#> 
#> $n_core
#> [1] 0
#> 
#> $n_border
#> [1] 0
#> 
#> $core_indices
#> integer(0)
#> 
#> $cluster_sizes
#> integer(0)
#> 
#> $is_core
#> [1] FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE
#> 
#> $neighbor_counts
#> [1] 1 1 1 1 1 1 1 1
#> 
#> $eps
#> [1] 0.5
#> 
#> $min_samples
#> [1] 5
#> 
#> $metric
#> [1] "euclidean"
#> 
#> $estimate
#> [1] 0
#> 
#> $n
#> [1] 8
#> 
#> $method
#> [1] "DBSCAN by BFS over density-connected cores; core detection delegated to grdbs"
#>